During the academic year 1978-79, school teams implemented a newly developed early identification system in all kindergarten and grade one classes in London, Ontario schools. After analysis and revision of the system, the internal consistency and concurrent validity of the process and a test of its short-term predictive validity were investigated. During the academic year 1979-80, the revised early identification system was applied in all kindergarten, grade one, and grade two classes. Results of this application were used to assess the internal consistency and concurrent validity of the system over a two-year period. School team ratings for the years 1978-79 and 1979-80 for those students who were in kindergarten or grade one during 1978-79 were compared. On the basis of the degree of agreement between the school team ratings and (1) the speech and language teachers' screening results, (2) the application of the resource allocation formula used with the Learning Resource System, and (3) the degree to which team ratings followed expected patterns with respect to students' grade, sex, and month of birth, it was concluded that the changes introduced have had relatively minor effects on the internal consistency and concurrent validity of the revised early identification system. (Author/MP)